AI Breaks the Way Consulting Charges

There is a sentence that gets said in consulting partner meetings, and it has been said for at least twenty years when I joined the industry.

We should be selling outcomes, not hours.

Everyone nods. Somebody says it is where the industry is going. Somebody else raises a practical objection about how you would actually measure it. The meeting moves on, and the next proposal that goes out the door is priced in hours.

I have been in that meeting. I have said the sentence. Nothing changed, and I want to be honest about why, because the usual explanations are wrong.

It is not that firms are greedy or unimaginative. It is not that clients refuse. It is something duller and much harder to fix, and ignoring it in the age of AI is expensive.

Why the hourly model actually survives

Firms bill by the hour because hours are provable and outcomes are not.

A timesheet is a fact. It is auditable, defensible in a procurement review, and it settles arguments. Especially when conusltants are onsite. When a client asks what they paid for, a firm can produce a record of who did what, for how long, at what rate. Nobody loves it. Everybody trusts it... well mostly.

Now try to bill on an outcome. You have promised a client that their program will deliver, say, two million dollars of value. Some of that is cost taken out, some is revenue accelerated, some is risk avoided. Your fee is a share of it.

Eighteen months later, who decides whether it happened?

The program changed shape twice. Two of the initiatives were canceled for reasons nobody could have predicted. The market moved. The sponsor who signed the business case has been promoted and her replacement has a different view of what success looked like. The finance team has numbers that do not reconcile to the ones in the original model, because the original model was built in a spreadsheet by a different firm consultant a year ago.

There is no argument you can win here, because there is no record either side agreed to keep.

That is the whole reason the industry still bills by the day. Not appetite. Provability. The billing model follows the measurement system, and the measurement system for outcomes has never existed in a form both sides would accept.

The thing that changes it is not what people expect

The common story about AI and consulting is a threat story. The tools get good enough to do the analysis, clients stop needing the firm, the work goes away.

I do not think that is the near-term problem. When I started Analyzt AI, that was my thesis. Could AI replace consultants. I don't think so, for many reasons that I'll cover in later blogs. The near-term problem is stranger, and it lands on the firms doing the best job of adopting AI.

AI compresses delivery time. Hourly pricing converts compressed delivery time into lost revenue.

Take an engagement I have scoped many times in one form or another. A business case for a significant transformation. On the low-end, starting around fifty thousand dollars, six weeks, two consultants part-time. On the high-end five to eight consultants, twelve weeks or more, and many hundreds of thousands of dollars.

Run the arithmetic. That is twelve to twenty four consultant weeks of effort. Which puts revenue somewhere between two and a half and five thousand dollars per consultant week, or roughly five hundred to a thousand dollars per consultant day.

If you know this industry, you already noticed that is low. That is well under list day rates for a firm of any standing. Which tells you something important about what that engagement really was. It was rarely a profit center. It was a door opener, priced to win the implementation work behind it.

Now compress it. The research, the financial modeling, the risk register, the scenario analysis, the first draft of the deliverable. A firm using its tools well can do that work in a fraction of the time. Call it a week and a half instead of six.

And here is the trap. The better you get at this, the less you are paid. You have taken a fifty thousand dollar engagement and turned it into a twelve thousand five hundred dollar engagement, a quarter of what it was, because your price was a function of your time and your time collapsed. The firms adopting AI hardest are cannibalizing themselves fastest, and they are doing it while congratulating themselves on the efficiency gain.

You cannot opt out either. Your competitor will compress, and their proposal will be cheaper and faster than yours.

Three doors, and two of them are walls

There are only three responses to this. It is worth being precise about them, because most firms are drifting into the first two without deciding to.

Door one. Pass the compression to the client. Charge twelve thousand five hundred instead of fifty. The client captures all of the value. Firm revenue on that engagement drops by three quarters. This only works if volume replaces the margin, which nobody has planned for and which I will come back to, because it is more interesting than it first sounds. Not many firms of a medium to large size even attempt this, they can't afford to.

Door two. Hold the price and pocket the compression. Still charge fifty thousand. Deliver in a week. Say nothing.

This is where a great many firms are quietly heading, and it is the worst of the three. Not because it is dishonest, though it drifts that way, but because it is fragile. It survives exactly as long as the client does not find out. And the client will find out, because their own teams are using the same tools and developing an intuition for how long this work takes now.

The moment that lands, the firm is not in a pricing conversation. It is in a trust conversation. And it will have arrived there by way of the exact hours based logic the industry says it wants to escape.

Door three. Decouple price from time entirely. Charge on the outcome. Compression stops being a threat, because your fee was never a function of how long it took.

Door three is the only one that survives contact with a client who understands the technology. It is also the one everybody in that meeting was nodding about twenty years ago.

What door three actually requires

Here is the part that gets skipped, and it is the reason the sentence in the partner meeting never became a practice.

Outcome pricing is only credible if the firm can prove the outcome later.

Not forecast it. Not model it in a deck at the point of sale. Prove it, afterwards, against what was promised, in a form the client's finance function will accept, over the life of a program that will run longer than the engagement that started it.

That is a measurement problem, and it is hard. It requires that the original business case is not a spreadsheet that leaves with the consultant who built it. It requires that the promised benefits are tracked as first class objects, with owners and dates and baselines, for years. It requires that when the program changes shape, and it will, the change is recorded against the original promise rather than quietly replacing it. It requires that a board can be shown, at any point, what was promised, what has been realized, and what the gap is.

Firms have not done this. Not because they did not think of it, but because doing it by hand is unaffordable. Nobody is going to staff a consultant for three years to maintain a benefits register for an engagement that ended eighteen months ago. So the record decays, the argument becomes unwinnable, and the firm goes back to billing days, because days are provable.

The measurement system has to be cheap enough to run for years after the engagement ends. That is the actual precondition, and it is a software problem rather than a consulting problem.

Which is why I think this changes now and did not change before. The same technology creating the pricing crisis also removes the reason the answer was unaffordable.

The better question

I want to come back to door one, because when I have put this to people who run firms, the conversation usually gets stuck on the wrong question.

The wrong question is what do we charge for a compressed engagement. Fifteen thousand? Twenty? People argue about the number and it goes nowhere, because it is a conversation about giving something up.

The right question is different, and I have never seen a firm answer it without their posture changing.

If a business case took one consultant week instead of twenty, how many would you run in a year?

Think about what actually limits that number today. It is not client demand. Every prospect a firm talks to could use one. It is not sales capacity. It is delivery capacity. Twelve to twenty four consultant weeks is a serious commitment of scarce senior people, so the firm rations them. It runs them for pursuits it thinks it will win. It says no to the rest.

Remember what that engagement is. A door opener. A pipeline generator for the margin carrying work behind it.

So the firm's constraint is not really its consultants. It is that its best pipeline generating instrument is too expensive to use very often.

Compress it, and the constraint moves. It stops being delivery capacity and becomes sales capacity, which is a far better problem, and one every firm already knows how to invest in. You are not running four of these a year for the clients you are confident about. You are running forty, including for the ones you were not sure about, which is precisely where the surprises live.

That is not a discount. That is a different business.

What this means for the deliverable itself

There is a second order effect worth calling out.

If the analysis takes a week rather than two months, the deliverable stops being an artifact and starts being a service. Today a business case is a document. It is correct on the day it is signed and begins decaying immediately, because refreshing it means paying for the work again.

When the marginal cost of a refresh approaches zero, there is no reason for it to be a document. It becomes a live position. The board sees what was promised, what has been realized, and what changed, this quarter and every quarter, without anyone rebuilding a model.

And notice what that gives the firm. Presence. The relationship stops ending when the deck is delivered. That is worth more than the engagement fee, and it is the thing that makes an outcome based fee defensible eighteen months later, because both sides have been watching the same numbers the whole way.

What we built, and why we built this part first

I have been describing a measurement system, so I should be direct about the fact that we build one.

Analyzt AI is a strategy execution platform. We started with strategy work deliberately, because it is where the promise gets made. A business case is a promise about the future with a number attached, and almost nothing in the industry is set up to hold anyone to it.

The parts that matter for what I have described:

Business cases are objects, not documents. A business case built in the platform has structured financials, assumptions, and predicted outcomes that persist. It does not leave when the consultant does, and it does not become a stale file in a folder.

Promised benefits are tracked against realized ones, continuously. Not as a report somebody runs, as a live position with owners, baselines and dates. This is the benefits realization layer, and it is the specific thing that makes an outcome based fee arguable later instead of unarguable.

The status pack rebuilds itself. Every period, from live data, at board altitude. That is what makes presence affordable, and presence is what outcome pricing depends on.

Fourteen production machine learning models score every strategy, outcome, initiative and business case. Predictive models, not generative ones. A confidence figure that is calibrated is a different thing from a confident sentence, and if a fee depends on a forecast, the difference matters.

It is methodology neutral. OKR, Balanced Scorecard, Hoshin, SWOT, Porter's Five Forces, Blue Ocean, PEST, McKinsey 7S. We do not have a house method and we are not going to sell you one. Your methodology is your product. It should not be ours.

The uncomfortable part

Selling outcomes has always been the answer that everyone agreed with and nobody implemented, because the thing standing in the way was never conviction. It was that you cannot charge for a result you have no way to prove.

That constraint is lifting. And the firms that move first will not be the ones with the best conviction about outcome pricing. They will be the ones that put the measurement in place before they need it, because you cannot retrofit a benefits record onto an engagement that finished last year.

The rest will spend the next few years doing door two. Holding the price, hiding the compression, hoping the client does not develop an instinct for how long this work takes now.

They will develop the instinct. They are using the same tools you are.

---

If you run a consulting firm and this is a live question for you, I would like to compare notes. We have a partner program, and the beta is open and free while it lasts.